YouTube Tests ‘Your Custom Feed’ to Improve Recommendations
YouTube is experimenting with a new tool designed to give users greater control over their video recommendations. The feature, dubbed “Your Custom Feed,” arrives as the platform seeks to solve common frustrations regarding its algorithm, which often struggles to accurately predict viewer intent based on past watch history.
Currently, users participating in the test will see the “Your Custom Feed” option displayed alongside the standard “Home” button on the platform’s interface. By selecting this tab, viewers can manually input specific prompts that define the types of content they wish to see, rather than relying solely on the platform’s automated suggestions.
Refining the Algorithmic Experience
The core objective of this test is to move away from passive consumption, where users are often fed content that does not align with their actual preferences. Historically, the algorithm has been criticized for over-indexing on singular topics—such as assuming a viewer is a lifelong fan of a franchise after watching only a few clips—resulting in a cluttered and irrelevant feed.
Key advantages of the new experimental feature include:
- Active Personalization: Users can input specific interests, such as “cooking” or other niche topics, to steer the algorithm toward more relevant video suggestions.
- Better Than Manual Filtering: The tool serves as a more efficient alternative to the manual “Not interested” or “Don’t recommend channel” options, which require individual interaction with every unwanted video.
- Direct User Input: Instead of waiting for the system to learn from history, users can proactively shape their viewing environment.
This initiative places YouTube within a broader industry trend of increasing user agency over content discovery. Other major platforms are exploring similar configurations to address algorithmic fatigue. For instance, Threads has been observed testing ways to adjust feed settings, while X is developing a feature that allows users to leverage its AI chatbot, Grok, to fine-tune their feed preferences.
While the long-term impact on user engagement remains to be seen, the introduction of experimental features like this suggests a shift toward more transparent and controllable discovery mechanisms across social video platforms.